Talk:Old Japanese

Untitled

I would like to expand this topic with much more detailed linguistic informtaion. I would appreciate help in merging it with the content already here. I will try to complete it within the next few days, depending on time constraints. Bendono 11:05, 1 October 2006 (UTC)

Questions

Am I correct in understanding that "Quadrigade" corresponds to the modern consonant-stem verbs, such as aruku, tatsu, etc., while "Upper Monograde" corresponds to modern vowel-stems? Do the "k-irregular", etc., mean that they end in ko₂, ki₁, ku, kuru, kure, ko₂? What is the difference between the two rows on each of the adjective types? E.g., what's the difference between -ki₁ and -karu? What are the differences between teh various pronouns listed, such as wa, a, ware, and are? Nik42 07:23, 16 October 2006 (UTC)

Thank you for the review. I added a few notes of clarification to the verbs and adjective sections. I will add more to the pronoun section a little later. The article is rather bare now and I plan on updating it in the near future. Bendono 08:53, 16 October 2006 (UTC)

Subscript numbers

Maybe I'm just missing it, but I don't see any explanation of what the distinction is between vowels with subscript 1s and vowels with subscript 2s. Kairos 21:55, 7 December 2006 (UTC)

It is discussed in the both the Phonemes and Transcription sections. Also, there is a link to Jōdai Tokushu Kanazukai. As stated, it applies to the entire syllable, not necessarily only the vowel. There are various theories, but there is no general academic agreement. In time it would be nice to discuss the theories in more detail. It is one of the most hotly debated topics in Japanese historical linguistics. If the wording is not sufficiently clear, please feel free to improve it or make suggestions as necessary. Bendono 00:30, 8 December 2006 (UTC)

Infobox

I believe it's somewhat misleading to claim that Old Japanese was spoken in "Japan", considering that the modern state is a rather modern concept. Hokkaido, for example, was colonized very late. I don't know the exact extent of the language but I feel the information should be more specific.

Do we have any information of the various stages of Japanese? When does the Old Japanese period end? Was it followed by "Middle Japanese" or did it go right into modern Japanese?

Peter ^Isotalo 20:48, 16 December 2006 (UTC)

Many wooden tablets ("mokkan") written and dated to the Old Japanese period have been excavated from all over the country. Also, several dialects are recognized. The most prominent of them is the "Azuma" (=Eastern) dialect which is contained in Books 14 and 20 of the Man'yōshū. (This is most definitely not Ainu.) Listing anything more specific than "Japan" will be most misleading and imprecise.

Old Japanese ends in 794 when the capital moves from Nara to Heian. It is succeeded by Late Old Japanese (sometimes called Classical Japanese). The general divisions are as follows: Old Japanese (-794), Late Old Japanese (aka Classical Japanese) (794-1184), Middle Japanese (1185-1333), Late Middle Japanese (C1333 - 1600) Early Modern Japanese (1600 or 1603 - 1867), Modern Japanese (1868-). Except for Modern Japanese, I plan on writing about all of these over time. Bendono 02:26, 17 December 2006 (UTC)

Very good info. I really look forward to seeing those articles.

Peter ^Isotalo 11:06, 17 December 2006 (UTC)

ISO 639-3

This article handles the same with ISO 639-3 "Old Japanese"? Since there is no example on the website, I am not sure. --Aphaia 01:45, 12 April 2007 (UTC)

The 639-3 description states: "A language of Japan. The ancestor of modern Japanese. 7th-10th centuries AD." There are several problems with this. First, Old Japanese lasts through the end of the 8th century. Late Old Japanese ranges between the 9th and 12th centuries. These two are rarely if ever grouped together, and even if they are, the century range does not correspond. Second, the ancestor of Modern Japanese is Middle Japanese, not Old Japanese. Middle Japanese may be broken into early and late middle Japanese. The registration is useless. Bendono 14:09, 12 April 2007 (UTC)

Thank you for your comment, I tried to find other historical Japanese including Middle Japanese and Late Old / Classical Japanese. I haven't found others and agree with you the registration is not helpful for us. --Aphaia 10:21, 13 April 2007 (UTC)

This is a v. useful and interesting discussion (thanks!), and deserves more prominence. I’ve added “ojp – but see caveat” in this edit, which explains the problem and links to this discussion.

—Nils von Barth (nbarth) (talk) 22:44, 15 August 2010 (UTC)

Explain why /r/ isn't initial?

So in this article we have a small section that talks how /r/ cannot be word initial. Wouldn't it be reasonable to explain why /r/ isn't initial? Something like this http://erssab.u-bordeaux3.fr/IMG/pdf/labrune_article_final_r.pdf could be used as source material. Of course, there are other theories as well that attempt to explain why /r/ isn't initial that can must be explained as well. So? FinalZero 17 03:03, 13 May 2007 (UTC)—The preceding unsigned comment was added by FinalZero 17 (talk • contribs) 03:02, 13 May 2007 (UTC).

See Kwamikagami's comment at Talk:Flap consonant#Initial flaps - difficult to pronounce?. --Florian Blaschke (talk) 16:15, 22 June 2015 (UTC)

Yeah those are all theories, but since there is no ironcald proof that Japanese is related to other languages, we can only speculate. In Taiwan there is a group of languages where it seems that /d/ became /r/ in all positions except the beginning of the word (or perhaps /r/ became /d/ word-initially), so that could be at work here, but again it's only a guess unless a miracle happsn and someone unearths a manuscript from 4600 BC written in proto-Koreo-Japanese. — Preceding unsigned comment added by Soap (talk • contribs) 17:32, 4 July 2015 (UTC)

下一段 verbs

I did some research on this while studying in Japan, and when we got to verb conjugations we read that there is only one verb that was conjugated in this manner, that being 蹴る. I've still got photocopies of the stuff somewhere I can dig up if anybody wishes to see them. Kaji01 07:49, 17 September 2007 (UTC)

No need. I have the entire OJ corpus. 蹴る is not attested in OJ. It first appears in LOJ (Heian period). Early examples can be found in 観智院本名義抄, 落窪物語, 栄花物語 etc. The OJ word for "to kick" was kuw- (蹴う) and it is 下二段. There are plenty of resources, several referenced, that clearly state that 下一段 does not exist yet at this stage of the language. Bendono 09:22, 17 September 2007 (UTC)

Whither e₁ and e₂?

In the phonological rules section, only six of the eight OJ vowels are mentioned. <a, u, o₁> in one class, <o₂> alone in another, and <i₁, i₂> being neutral. What of e₁ and e₂? If the i-variants hadn't been mentioned, one could assume they were neutral, but right now they're in a kind of limbo. Anyone care to correct this? --Wtrmute (talk) 13:25, 19 December 2007 (UTC)

It may be best to omit that detail. The rules were discovered by Arisaska Hideo in 1934. Regarding e₁ and e₂, he said that they "should be feminine [second group] or neuter"; however, in his data they are masculine (first group). Various other resources list slightly different groupings. For example, Ōno (2000) gives {a, i₁, u, o₂} and {e₁, e₂, i₂, o₁}. He also goes on to seriously discredit the vowel harmony theory. The section is weak, but the resources are not very consistent on the issue either. Bendono (talk) 14:28, 19 December 2007 (UTC)

Mixing old Japanese with Jōdai nihongo

This article is mixing old usage of Japanese with 上代日本語 (Jōdai nihongo). In Jōdai nihongo, much vocabulary of Yamatokotoba is found than Chinese Kango. We shouldn't mix borrowing Chinese characters with pronouncing the vocabulary in Yamatokotoba. Huge number of Japanese pronunciation exsisted in Jōdai nihongo. --210.139.234.58 (talk) 13:09, 21 September 2008 (UTC)

Category:Extinct languages vs. Category:Extinct languages of Asia

Category:Extinct languages of Asia is itself a category within Category:Extinct languages. — Robert Greer (talk) 09:03, 14 August 2009 (UTC)

Why are Old Japanese and Old Chinese classified as 'extinct'? I don't see people classify Old English as 'extinct'--Tricia Takanawa (talk) 16:48, 8 October 2009 (UTC)

attn Bendono: the meaning of K

I'm afraid that, if K means /k/, as was indeed my first reading, that paragraph's argument is incoherent to me. What does it even mean to distinguish /ko₁/ and /ko₂/ "except for /wo/"? By contrast, if K is a cover symbol for consonants, the reasoning is along the believable line that /o/ and /wo/ might only fail to be distinguished after /w/. And anyway, if K did mean /k/, why didn't the original writer set it in lowercase? Do you have a good reason (e.g. a source!) for your interpretation? 4pq1injbok (talk) 05:58, 7 July 2010 (UTC)

My apologies, I did not read your edit very well. Being the original editor, I had written it as /Co₁, Co₂/. I must have missed when someone later changed it to k. I have previously made this same correction before, so had probably been confused by your edit which I had assumed was still correct. Here is the source of the mistake. I'll restore it. Regards, Bendono (talk) 06:20, 7 July 2010 (UTC)

Thanks.

---

If I might append a different question about the 1 / 2 distinctions here. What's the actual positive evidence for the lack of the distinction in those cases where it's not recorded? Do we know, say, that whichever form of Chinese supplied the man'yougana would have been capable of writing distinctions like /si₁/ vs. /si₂/ if they had existed? 4pq1injbok (talk) 08:37, 7 July 2010 (UTC)

Good question. If there was distinction between samtōji 三等字 the third grade character and yontōji 四等字 the fourth grade character in 歯音 /ts,tʃ,dz,dʒ,s,ʃ,ʒ,z/ group of Chinese phonetics, there would be /si₂/ vs. /si₁/ and /zi₂/ vs. /zi₁/. For example, in Old - Middle Chinese If there was /sɪǝg - sɪei/, /sɪar - sɪě/, /sɪuǝd - sɪuǝi/, /thɪǝg - tʃɪei/, /thɪar - tʃɪě/ or /thɪuǝd - tʃɪuǝi/ development, it should be /si₂/. However, that development never have done. More simplly, if there was a pronunciation like /sɪi/, /sɪui/, /tʃɪi/, /tʃɪui/, /tsɪi/ or /tsɪui/, it should be /si₂/. The difference between Samtōji and Yontōji was only Kaion 介音 /ɪ/ and /i/. Type 1 was /i/, type 2 was /ɪ/. Thank you. ஒலிசிந்நெக் (talk) 17:22, 26 November 2010 (UTC)

Good news. di₂ and ri₂ are probably found, said Wikiresearch:Old Japanese. I am also sure that it's positive. ஒலிசிந்நெக் (talk) 05:57, 28 November 2010 (UTC)

What is the southern Japan ?

In phonology section. There are northern(north-eastern), eastern, western and most-western Japan.

However, I don't know southern Japan, esp. in the Old age.

West-southern Japan is Ryūkyū.

If it means Kyūshū, esp. the Dazaifu, it should be written most-western Japan.

And in the Old age. Japanese is not so different from most-western to eastern.

For example, to separate Pannatu-Pannatsu-Pangatsu are only consonant gradation of -nn- and -ng-. a house •ipë-•ipa are only existance of concrete postfix -i or not. *•ipa-i > •ipë. to where ni-na are conjugation as nna-nni-nnö (and nga-ngi-ngö).

ஒலிசிந்நெக் (talk) 06:13, 24 November 2010 (UTC)

Please improve my broken English.

I wrote facts, but not elegant nor fluent English. ஒலிசிந்நெக் (talk) 07:53, 24 November 2010 (UTC)

Whether Old Chinese had open syllables

Thank you for your qualifying, 4pq1injbok.

You wrote in your comment "...depends on your reconstruction" but reconstruction is not mine but of Toudou Akiyasu(藤堂明保). He is No.1 scholar of Old Chinese phonetic study in Japan. Why was not cited more of his great works in English Wikipedia? ஒலிசிந்நெக் (talk) 12:19, 26 November 2010 (UTC)

You're welcome. I don't know why, aside from the generic reason that most Anglophones can't read the Japanese literature. I see we don't even have a page on Toudou Akiyasu at present. Perhaps you'd like to start one?

I haven't seen Toudou's reconstruction, myself. I'm no expert on any of this, so take this only for what it's worth: but I find it unlikely on typological grounds that a language of the area and time would have no open syllables; and I think Baxter's solution to the problem of rùshēng-nonrùshēng contacts of reconstructing *-ps *-ts *-ks is preferable to Karlgren's old idea of *-b *-d *-g, on which the closed-syllable-only reconstructions I know of are predicated. 4pq1injbok (talk) 20:20, 29 November 2010 (UTC)

Hi. rùshēng-nonrùshēng (入声-不入声) dualism is now obsoleted. yánglèi-rùlèi-yīnlèi (陽類-入類-陰類) triangle theory is instead in Old Chinese (上古漢語). rùshēng are only codas /p, t, k/. Codas /g-k-ŋ/, /r-d,t-n/, /Φ-p-m/ triangle pairs are reconstructed. Note that Φ is not fai but zero(not exist). Coda /d/ became qùshēng (去声) in Middle Chinese (中古漢語). Thank you. ஒலிசிந்நெக் (talk) 16:02, 30 November 2010 (UTC)

"Writing system" section

The first part of the "writing system" section needs serious attention, particularly this part:

"On the other hand, Chinese characters are used semantically to represent Japanese syllables. It is called kuniyomi or kun'yomi. For example, 猿(monkey) was read as maci (not saru) as a particle because monkey was called maci(ra), not meaning monkey but phonetically maci (recently it is called ate-ji). It became more prevalent. For example, 八十一(81) was read as kuku(9x9) and 十六(16) was cici(4x4)."

First of all, there is no reason to use [c] in the transcription; it is not used anywhere else in the article. [s] is fine.

Secondly, because /masi/ is completely extinct in modern Japanese, only a very few people who already understand the topic will understand what is intended. 谷 used for the particle /dani/　would at least be tenuously connected to the modern Japanese people know.

Thirdly, this example appears before a clear statement that 猿 could be used to represent /masi(ra)/ as in monkey, although I suppose this is implied in the reference to kanbun above; still, if you weren't already with the intricacies and history of the Japanese writing system I think you would be very hard pressed to understand what is going on.

I propose that this system be rewritten to show how the use of characters spread more naturally, rather than leaping right to the extreme cases:

1) Use of Chinese chars to represent Chinese (semantic and phonetic relation to chars): 谷 for the current Chinese pronunciation of 谷 2) Use of Chinese chars to represent Japanese words (semantic but no phonetic relation to chars): 谷 for /tani/, "valley" 3) Use of Chinese chars to represent Japanese sounds (phonetic but no semantic relation to chars): 与 for /yo/, etc. 4) Use of Chinese chars to represent Japanese words (no semantic or direct phonetic relation to chars, but secondary phonetic relation via 2 above exists): 谷 for /dani/

I know that the standard Wikipedia response is "go ahead and do it, then," but I do not have the academic expertise to say which of the above preceded what, how the evolution happened, etc.

However I note that several contributors to this talk page clearly do, and I urge them to restructure this section to make it clearer even to people who have not learned any Old Japanese yet (a test the 猿 example fails). —Preceding unsigned comment added by 210.80.252.2 (talk) 05:26, 29 November 2010 (UTC)

Thank you for your comment and attention. First of all, yes, I am not good at English. Althought I AM AN EXPERT OF OLD JAPANESE. Old Japanese uses phonetically kun'yomi even it is not related with its meaning. For example, to write 猿 is equal to write まし in hiragana. It is similar to that to write 猿 is equal to write masi, if early middle Japanese people knew alphabet. See you later and again. ஒலிசிந்நெக் (talk) 15:36, 29 November 2010 (UTC)

Hello ஒலிசிந்நெக்; yes, I see from your edits that you are indeed an expert on Old Japanese. I don't dispute that 猿 was used to write /masi/ (in contexts other than meaning "monkey" (/masira/). My point is that since the word /masira/ meaning "monkey" is now extinct, it is a confusing example. (Although a cool one.) I think that since /tani/ 谷 is still a living word, it will be more relateable for people. Also, we need a statement using Chinese characters to represent Japanese words based on semantics predated the second-stage use of Chinese characters to represent Japanese word groups unrelated by semantics. I will go ahead and make some changes, then; please feel free to edit it if I am wrong in the particulars! —Preceding unsigned comment added by 219.206.98.238 (talk) 21:54, 29 November 2010 (UTC)

[this space reserved]

I'm back. First of yours, you are right. /s/ is fine. I wrote those paragraph before "/s/ = [s, ʃ, t͡ʃ, t͡s, c]".

Second of yours, you are better. 谷 /dani/ is attested ex. 万01/0018.

Third of yours, it is not probed yet that kun'yomi predates kun'gana (see below). Sophisticated those days people could write Chinese character(i.e., kanji). Then, could they all write Chinese sentences (i.e., kanbun)? No, they ALL could NOT. Only some smart people could write kanbun.

There were six ways to write Old Japanese using Kanji.

1. On'gana: one character represents one syllable phonetically
2. Kun'gana: one character represents one or more syllables phonetically
3. Puzzled Kun'gana: sort of 八十一 /kuku/, 十六 /sisi/
4. On'yomi: one character represents one syllable semantically (a few instances)
5. Kun'yomi: one character represents one or more syllables semantically
6. Kun'doku: Reading kanbun as Japanese (just in time translation)

Those facts should reflect to main article. Thank you. ஒலிசிந்நெக் (talk) 05:51, 30 November 2010 (UTC)

References

I spent a while time looking for "Jōdai Izen no Wago", but came up with nothing. There wasn't anything by that name in WorldCat. If it is to be included, it needs to be properly cited.

There were several other issues:

• Is the publisher Shinchō-sha?
• If it's in a journal, which is the template it was using, what's the journal's name?
• What is the ISBN? The one given was not a valid ISBN of any sort.

I have removed it because of these issues and because it was not cited in the body of the article. It would be nice to have more sources, either in Japanese or English, but it should be done right.

Otherwise, I've cleaned up the references so they are consistent (only romanized Japanese), put Frellesvig's stuff into the proper template, and added Vovin's grammar. --Limetom 06:11, 20 October 2011 (UTC)

Proto-Japanese: Issues and Prospects  By Bjarke Frellesvig, John Whitman

http://books.google.com/books?id=aun8BRHTDEAC&printsec=frontcover&source=gbs_ge_summary_r&cad=0#v=onepage&q&f=false

Rajmaan (talk) 04:34, 16 December 2012 (UTC)

derivation of e2, o2

I can't access all of Miyake. Can someone add the derivation of e2 and o2 from pOJ? — kwami (talk) 23:19, 17 April 2014 (UTC)

Yamato is "old Japanese"

I find this entire page to be entirely lacking and I wonder why distinct language types - such as Yamato are not actually given a section or overview and instead lumped into "Old Japanese". Is that even correct I wonder? Anyone know for sure? ChrisGualtieri (talk) 15:13, 31 July 2014 (UTC)

Notation inconsistency

It appears to be a contradiction throughout the article when using the vowels o₁ and o₂: according to Transcription section, o₂ is ö and the Vowels one states that o₁ was /ə/ (or rather [ə]?) and o₂ (ö) was /o/ according to Miyake. Also, *ə was included in the PJ inventory.

However, Proto-Japanese says 'the proto vowel system can be reconstructed as /*a, *i, *u, *o₂/' , thus matching o₂ with *ə, in contradiction with the sections above. Also, Wiktionary uses o₂ for *ə (Wiktionary: 心).

Moreover, Miyaki does not use subscript notation. A possible solution to this problem is to match o with o₂ to make clear that it was o₂ the primary vowel rather than o₁. If we were to retain the matching o = o₁, the Proto-Japanese section should be modified. --Farru ES (talk) 18:24, 28 December 2014 (UTC)

Yeah the schwa is definitely o1. I thought I fixed that already but apparently not so thank you for noticing. I will fix it later. Lollipop (talk) 23:39, 27 June 2015 (UTC)

Fix,ed although the verb conjugs may still be wreong.—Soap— 17:30, 4 July 2015 (UTC)

Jōdai Tokushu Kanazukai

Existing text: It distinguished between a few pairs of syllables with identical pronunciations — a phenomenon known as Jōdai Tokushu Kanazukai

• Firstly a small point: "between a few pairs" of course means between the members of a few pairs; my view is that this point is not too 'picky' for a minor edit, and my intention is to do one soon.
• Secondly a question: doesn't "identical" here mean apparently identical, or some such?

Peter010101 (talk) 05:19, 23 November 2015 (UTC)

Erased

@Kanguole:, your changes erased other people's sourced work, you should at least justify it in the Talk Page. 2A01:E35:8A85:7E40:41FB:7335:1AF2:6FE1 (talk) 13:01, 5 June 2017 (UTC)

Your addition cited Frellesvig's draft chapter on the Cornell website; I changed that to the published version (page 32 of his book). Your text "since Lange, it has been the glide theory which is more generally accepted" has a quite different meaning from the Frellesvig's "since Lange (1973), it is common to interpret the difference in terms of sequential diphthongs". That is, "common" does not imply "more generally accepted". Kanguole 13:41, 5 June 2017 (UTC)

Please answer what my comment said. I was not talking about, nor do I care about, my own modifications. The fact you felt the need to justify deleting my edits says a lot about your mindset. 2A01:E35:8A85:7E40:41FB:7335:1AF2:6FE1 (talk) 19:20, 26 June 2017 (UTC)

I have edited per WP:BOLD, with edit summaries describing each edit, re-organizing the text to avoid repetition, and basing it on solid secondary sources. If any specific edits are of concern, let's discuss them. Kanguole 14:49, 27 June 2017 (UTC)

Derivation of yamaji

In the Pre-Old Japanese section, Old Japanese yamadi is described as arising from yama + mi₁ti. This is problematic for a few reasons.

• The mi on the front is an honorific prefix. The core term is just ti, modern 路 (chi).
• This mi prefix appears to resist contraction in cognates in the Ryukyuan branch, suggesting it would likewise resist contraction in the mainland Japanese branch.
• The term yamamiti appears in the Man'yōshū as a term distinct from yamadi.

Consequently, it seems more likely that yamadi arose from yama + no (possessive particle) + ti → yama no ti → yamanti → yamadi.

The first sentence in the section is cited to Miyake and Frellesvig. From what I can glean of Miyake's text in Google Books, he doesn't address the derivation of yamadi. I don't have access to Frellesvig's work, so I cannot confirm if he explicitly describes yamadi as from yama + mi₁ti, and whether he addresses the issues I mention above. ‑‑ Eiríkr Útlendi │^{Tala við mig} 00:52, 18 September 2020 (UTC)

This Cornell page has parts of Frellesvig's book (with less pretty formatting) as handouts, including chapter 2, The Phonology of OJ. These examples are on page 22 of that handout. The derivations aren't discussed further.

Co-existence of multiple forms of a word can actually be evidence in favour of a contraction, but note that all the instances of yamamiti (and some of yamadi) in the Man'yōshū are written logographically, as 山道. Kanguole 09:19, 18 September 2020 (UTC)

FWIW, Alexander Vovin also derives yamadi from yama no ti, as shown on slide 10 of his presentation 『上代日本語と古代・中世韓国語の「水」と「涙」』 available here at Academia.edu (free registration might be required). ‑‑ Eiríkr Útlendi │^{Tala við mig} 18:50, 17 December 2020 (UTC)

Removed flawed example of vowel elision from the Old Japanese#Morphophonemics section

I've removed the following line:

• to₂ko₂ + ipa → to₂ki₁pa

The relevant section is attempting to show cases where [PART 1] + [PART 2] results in the elision of the initial vowel of [PART 2]. However, as we can see here, [PART 1] changes, such that to₂ko₂ becomes to₂ki₁. Consequently, this is clearly not just an example of vowel elision, making this inappropriate for this section.

‑‑ Eiríkr Útlendi │^{Tala við mig} 19:15, 4 August 2021 (UTC)

The section illustrates a range of treatments of adjacent vowels. This particular example was cited as an illustration of elision of the first of two adjacent vowels, i.e. the final vowel of [PART 1], and that seems to be exactly what it shows (i becomes i₁ because the distinction occurs after k). Kanguole 19:42, 4 August 2021 (UTC)

@Kanguole: Given the description, "In other environments, the first vowel was dropped", I think the main issue is the wording of that description -- I interpreted "first vowel" here to mean the "first vowel" of ipa, as I tried to explain above ("...results in the elision of the initial vowel of [PART 2].").

I'll have a go at rewording that description for better clarity, in light of your reply. ‑‑ Eiríkr Útlendi │^{Tala við mig} 19:53, 5 August 2021 (UTC)

There are not four patterns but three: the other form of 'my house' is considered an instance of the second pattern. Kanguole 20:57, 5 August 2021 (UTC)

@Kanguole: The vowels behave as in #2, but the overall phenomenon is more than that, since it involves fusion of what had been independent lexemes. I added a comment to that effect, that the outcome is essentially #2. If you'd prefer to restructure the section, please do so. ‑‑ Eiríkr Útlendi │^{Tala við mig} 23:56, 5 August 2021 (UTC)

No noun grammar information

Someone should make a section describing the grammar of nouns in Old Japanese Mizoru (talk) 19:05, 6 September 2021 (UTC)

@Mizoru: I'm curious, what would you want to see? Nouns and other non-inflecting words generally don't have grammar. Are you asking about how particles are used? ‑‑ Eiríkr Útlendi │^{Tala við mig} 00:15, 27 October 2021 (UTC)

Old Japanese#Inflected forms section

@C44986054 and Kanguole: You two both recently worked on this section. I'm curious about the specificity of the note there about "[this six-form system] also fails to capture ku-inflection and mi-inflection."

I'm familiar with ク語法, a specific kind of nominalization with some wonky ways of attaching, as described in the entries at Kotobank for ク語法 and く (as a suffix).

I'm not familiar with the ミ語法 added in this edit, and there is no such entry at Kotobank that I can find. Is this in reference to the み suffix found on adjectives? If so, this doesn't seem to belong in a section describing verb inflections. If this is in reference instead to the み attaching to verbs in a non-exhaustive listing, this attaches to the 連用形 in a straightforward manner in line with normal rules for compounding verbs, and thus this does indeed seem to be adequately explained by the six-form inflection system.

As it is, I find myself thinking that the previous text here was better: "It also fails to capture some inflected forms." That's pretty clear without being overly specific. What value is there in changing this text? ‑‑ Eiríkr Útlendi │^{Tala við mig} 19:26, 25 May 2022 (UTC)

Indeed the addition isn't very informative. It is citable, though: Frellesvig p117 gives examples not accounted for by the six forms: stative -yer-, nominal -(a)ku, exclamatory-1 -sa and infinitive-2 -mi.

Also, the contrast claimed under Mizenkei between "English-speaking" and "native Japanese" scholars is unfounded. Kanguole 22:15, 25 May 2022 (UTC)

@Kanguole: Cheers, thanks for the reply. I have no argument with the statement that this six-form model "fails to capture some inflected forms" -- I've seen enough oddness in how some of the more-obscure auxiliaries attach to know that there is sometimes more going on than the six-form model can easily explain (at least, at the surface level).

My concern was with the rewording, which makes it sound like the only inflected forms that cannot be explained are the ku-inflection and mi-inflection. I'll rework that particular sentence for now. ‑‑ Eiríkr Útlendi │^{Tala við mig} 22:29, 25 May 2022 (UTC)

@Eirikr:It's my fault to ignore they're just combination of these 6 stems with "み" and "く". @Kanguole: "There's a assumption...." would be better.--C44986054 (talk) 04:31, 26 May 2022 (UTC)

• @Kanguole: Re: the mizenkei, what about -ba? That doesn't include any initial a- element, which would seem to poke a hole in the definitive statement attributed to Frellesvig.

I also note the existence of a few rare nouns that appear to be derived from verbs and end in -a -- at the moment, only 塚 (tsuka, "mound") comes to mind, ostensibly from 築く (tsuku, "to build up"). ‑‑ Eiríkr Útlendi │^{Tala við mig} 00:23, 27 May 2022 (UTC)

I believe Frellesvig is talking about different periods, with auxiliaries starting with *a- in the pre-OJ period, the a being re-analysed as part of the stem in the OJ period and other auxiliaries following by analogy.

Regarding nouns, I suppose there's also wosa 'chief' vs wos- 'to rule', but many explanations are possible, and we'd need a source saying it anyway. Kanguole 17:01, 27 May 2022 (UTC)

But there must be very few evidence about pre-OJ( since the Japanese artifacts in this period are mainly Chinese but not the native Japanese texts) .From this point of view, even if Frellesvig did talk about pre-OJ, this assumption is too weak to replace the theory of mizenkei. C44986054 (talk) 13:33, 29 May 2022 (UTC)

Pre-OJ is based not on direct evidence, but on internal reconstruction applied to OJ. Mizenkei is an analysis of OJ, which is a different period. Kanguole 14:26, 29 May 2022 (UTC)

Additionally, I think accepting a "C+V" closed form is not disturbing. Just like Japanese is pitch accent while English is stress accent. A vowel or a consonant is not necessarily a single units in mind for every language (As far as I know there is not a concrete theory to support these kind of assumption). In my opinion, accepting this kind of difference is important for diligently studying a foreign language (Therefore, I think kana is more better than romanization.) C44986054 (talk) 13:49, 29 May 2022 (UTC)

Many of the processes one wants to discuss here, including inflection of verbs, fusion of vowels, origin of voiced obstruents, loss of sounds, etc, operate on a vowel or consonant mostly independently of the rest of the syllable. They are thus much more transparent in a phoneme-based presentation than in a syllable-based one, and that is why so much linguistic work uses romanization. Kanguole 14:26, 29 May 2022 (UTC)

There's two problems:

(1)Latin alphabet is ambiguous, because it hints the reader to pronounce in a "western-language" style (but not the standard IPA) like ti as [ti] but not [t͡ɕi] or si as [si] but not [ɕi]. (especially for English native speakers, due to the irregular spelling of the vowels and the different pronunciation of the vowels from Japanese)

(2)We need evidence to demonstrate the roots with a divided vowel or consonant was really a matter in the mind of old Japanese people. Otherwise, it's just a convenient explaination. The phonetic division of Mizenkei is consistent with all the stages of Japanese with handwriting records. But Frellesvig's theory acquires some "ad hoc" rules like /i/ + /aba/ -> /iba/ or /e/ + /amu/->/emu/.

As for learning beginners of Japanese, romanization is friendly but just a temporary mean before they completely memorizing the whole kana.

C44986054 (talk) 15:47, 29 May 2022 (UTC)

Let's focus on the issue of romanization vs kana, and leave Mizenkei to one side for the moment.

If we look at say the conjugation of the exemplar yodan verb kaku 'write', we see the common part is kak- with different vowels added for the different stems, and we can apply that to any other verb in the class. And similarly for the other verb classes. The natural analysis splits between the consonant and the vowel. This is transparent when we look at the sequence of phonemes in the word, but obscured if we use kana. You can talk about columns and rows of the kana table, but in doing so you are analysing the syllable into consonant and vowel.

Similarly, the developments discussed in the "Pre-Old Japanese" section often treat the consonants and vowels separately, which is straightforward with a phoneme-level transcription. It's not a matter of reading people's minds, but rather economy of analysis, a common goal in linguistics.

The default interpretation of letters in linguistic articles follows the IPA, but it should be clear here that a broad phonemic transcription is being used. You seem to assume that modern Japanese pronunciation is the best guide to Old Japanese pronunciation, but this is far from certain, and indeed known to be false in several instances. Kanguole 17:38, 29 May 2022 (UTC)

What is your fundamental assumption of your "natural analysis"? "Transparent" and "obscured" are just the point of view from languages with separate vowels and consonants like English (i.e. very convenient). And this kind of analysis surely implies Pre-Old Japanese was very different from the later Japanese. There's also an other example about stereotype of linguistics. Before European linguist studied Sanskrit, they didn't know sonorants can also be the kernel of a syllable.

Although Inflection of Japanese just replaces consonants, japanese people still can not spell the end consonant of Chinese characters of Checked tone without the "closed" Kana.(ex. 學 Middle_Chinese [1] /ɣok/ but カク in 呉音[2]) Therefore, this kind of "closed" mora may not be the influence from middle Chinese, and possibly be native.

We need concrete evidence to proof that old people did treat them independently. Otherwise it's just a kind of explanation, but not the fact. As for IPA, advanced (and a little nitpicky) concept of modern linguists doesn't mean old people should think such carefully (ex. classical latin grammarians can't distinguish between aspirated and unaspirated consonants)

And you can't leave Mizenkei to one side. Conversely, you have to agree that Frellesvig's explanation is not surely true, so this kind of division is controversy until a miracle handwriting record tell you old people did split it down QAQ C44986054 (talk) 18:57, 29 May 2022 (UTC)

• @C44986054: Your concerns touch on two separate issues -- graphical representation (kana vs. romanization) is one issue, and the validity of the mizenkei is another issue. These are separate concerns. Kanguole's comment about "leav[ing] Mizenkei to one side for the moment" is a clear indication to me that Kanguole wished to discuss graphical representation first, separately from the unrelated issue of the mizenkei.

That aside, I confess I do not understand much of your latest response. As best I can tell, it appears that you have misunderstood much of what Kanguole was saying.

About borrowings and additional syllables to mimic closing coda consonants, Hawaiian and Māori both do something similar, where borrowed English terms that have closing coda consonants are represented by consonant + vowel combinations in the Hawaiian or Māori. English CVC thus becomes Hawaiian or Māori CVCV as a natural consequence of the latter languages not allowing coda consonants. Consider English sheep and Māori hipi, or English Christmas and Hawaiian Kalikimaka, among numerous examples. ‑‑ Eiríkr Útlendi │^{Tala við mig} 21:55, 31 May 2022 (UTC)

Surely, separate vowels and consonant will make graphical representation clearer, but we should keep in mind that these arrangement is bases on modern linguistic concept. For example, there was not "u" in classical Latin(but only "v"), because romans didn't distinguish /w/ between /u/. v with /u/ is always written as "u" in modern orthography, while v with /w/ is written as "v". Old fashion (萬葉假名) is ambiguous but faithfully reflects how old people spelled these sound. Similarly, in modern Japanese, kana and romanization have their own advantages in "graphical representation", but I prefer to native writing system. C44986054 (talk) 16:34, 1 June 2022 (UTC)

Question about final comment about bigrade verbs from adjectives in the Old_Japanese#Conjugation_classes section

Specifically about this part:

Some bigrade bases also appear to reflect pre-Old-Japanese adjectives with vowel stems combined with an inchoative *-i suffix:

• *-a-i > -e₂, e.g. ake₂- 'redden, lighten' vs aka 'red'.
• *-u-i > -i₂, e.g. sabi₂- 'get desolate, fade' vs sabu- 'lonely'.
• *-ə-i > -i₂, e.g. opi₂- 'get big, grow' vs opo- 'big'.

• The first one might ostensibly be about verb 明ける (akeru, "to brighten, to become bright"). However, I am more accustomed to seeing this described in relation to 明るい (akarui, "bright") than 赤い (akai, "red"), and I am not aware of any "to redden" sense that is expressed with any verb akeru. Moreover, the verb root ak- seems much more closely tied to ideas of "to open" than "redden".

• The second one works well enough: Old Japanese adjective 淋し (sabushi), source of later modern form 淋しい (sabishii, "lonely; dissatisfied, missing something; desolate, shabby, rusty"), and bigrade verb sabu → sabi- ("to become lonely, dissatisfied, desolate").

• The third one loses me. Considering that medial -p- in Old Japanese later lenited, any such verb stem opi- in Old Japanese would manifest in the modern language as oi-. The closest verb I can find that might match is 老いる (oiru, "to grow old"). However, this has nothing to do with growing bigger, and the bigrade conjugation derives from older form oyu, not opu or opo. Searching dictionaries that include Classical Japanese terms for any verbs beginning with おひ- (ohi-) or verbs with conclusive form おふ (ofu) also nets me nothing relevant. I cannot tell if that is the fault of the references, however, in possibly omitting obscure and obsolete terms.

I have checked the Unger resource listed on JSTOR, and the indicated page 665 includes nothing about these three specific examples ("redden; get desolate; get big"). I don't have access to the other two resources, Frellesvig or Whitman.

Could someone double-check the first and third examples and help explain what the authors were describing? ‑‑ Eiríkr Útlendi │^{Tala við mig} 00:40, 5 August 2022 (UTC)

The examples are in both Whitman and Frellesvig. Whitman glosses aku/ake₂- as 'dawn, redden' (homophonous with the verb 'open'); Frellesvig has 'redden, lighten'. Whitman glosses opu/opi₂- as 'grow'; Frellesvig has 'get big, grow'.

Elsewhere, Frellesvig has aka- 'bright, red', aku/ake₂- 'become bright' (p91), aku/ake₂- 'dawn' (p5) and opu/opi₂- 'grow, get bigger' (p98).

Frellesvig (p207) has OJ opi₂ > EMJ opi > owi > LMJ oi 'grow.INF'. This is apparently 生ふ (ou). Kanguole 14:58, 5 August 2022 (UTC)

@Kanguole: Aha, very helpful! Now with the Japanese spelling I have a better target to search for. Much appreciated! This 生ふ (ou) does appear to be obsolete in modern JA, so not included in many references. I was surprised that I was having trouble finding it in the Nihon Kokugo Daijiten since that is usually so good about listing even obsolete terms. But then I've also learned that half the challenge with the NKD is sussing out how they've lemmatized and indexed their entries, and what exactly you need to search for to find something -- and this is another such case. Now with the spelling, I am finding it.

I also appreciate your addition of further detail from the references. I've long wondered if there might be any connection between the "open" and "red" senses in apparent etymon root ak-, perhaps via "open" → "begin" → "dawn" → "color of dawn: red", but from what you've got here, it sounds like Frellesvig at least views this as an accidental overlap in phonology. I could similarly waste time trying to rationalize English ball (round bouncy thing) and ball (big formal dance). 😄

Thanks again! ‑‑ Eiríkr Útlendi │^{Tala við mig} 16:41, 5 August 2022 (UTC)

To be clear, Frellesvig doesn't compare the two verbs; he talks about both of them in different places and consistently labels them with separate meanings. Kanguole 17:03, 5 August 2022 (UTC)

@Kanguole: Good to know, thanks for the clarification. The implication would seem to be that he views these as separate etyma, despite not saying so explicitly. I hope some day to get my hands on his History of the Japanese Language (and ideally time to actually read it thoroughly). Cheers, ‑‑ Eiríkr Útlendi │^{Tala við mig} 22:39, 5 August 2022 (UTC)

What happens in the third is that the I is not a nominative particle in this case, but a verbalizing particle, which in this case would be a kami nidan katsuyou.

compare with oyu (to age) -> oya (parents) + i

===sources===

https://oncoj.orinst.ox.ac.uk/cgi-bin/oncoj_dictionary.sh?search=L030407a&part=n

https://oncoj.orinst.ox.ac.uk/cgi-bin/oncoj_dictionary.sh?search=f000053&part=n&db=oncoj The Young Prussian (talk) 01:57, 22 December 2022 (UTC)

The article needs to use examples that are worked out by sources, not based on our own analysis of data like the ONCOJ dictionary. Kanguole 18:08, 22 December 2022 (UTC)

This is not my data analysis, and the ONCOJ dictionary is no longer a source? The Young Prussian (talk) 19:19, 1 February 2023 (UTC)

I thought you were suggesting a different analysis of the third example quoted from the article above, but I must admit I'm not sure what you're getting at, or why oyu/oya is relevant. Kanguole 23:16, 1 February 2023 (UTC)

There is nothing that contradicts the previous examples, it is just one more form of contraction of vowels that had not been cited in any other example on the page, so of course it is relevant, and there is no way, somehow, this contradicts the third example or any other, and more verbs that are formed by the same variation, like koi- (lie down). The Young Prussian (talk) 00:55, 2 February 2023 (UTC)

• In an attempt at clarity -- oya is related to oyu, per the ONCOJ link, but that entry does not make the claim that oya + i = oyi.

There is a small group of nouns ending in -a that appear to be related to verbs. Terms like tsuka "grip" and tsuku "to put one's hand to something", or related tsuka "mound" and tsuku "to build up", or nawa "rope" and nau "to plait or braid cordage", or possibly even shima "island" and shimu "to close off". For these, tsuku and nau are both from yodan verb paradigms, which appears to rule out any need for the -i fusion discussed above. Meanwhile, oyu is only attested as kami nidan in ancient sources, and shimu as shimo nidan (with the "close" sense). The exact nature of the relationship between these verbs and nouns isn't clear: is one derived from the other, or are both derived from a root? Might depend on analysis approach. ‑‑ Eiríkr Útlendi │^{Tala við mig} 02:22, 2 February 2023 (UTC)

Negative, the entry explicitly shows "oya" + "i" as a derivation of oyu, compare with sabi- ([3]), it's correct. The Young Prussian (talk) 02:34, 2 February 2023 (UTC)

• I don't think you're reading these entries correctly. ‑‑ Eiríkr Útlendi │^{Tala við mig} 04:21, 2 February 2023 (UTC)

  I'm sure it does, because the dictionary puts the same derivation for both sabi- and oi-, so it doesn't make sense for oi- not to be included here. The Young Prussian (talk) 15:54, 2 February 2023 (UTC)

Again, I don't think you're reading these entries correctly. You also appear to be confused in your reply here?

To sum up:

• The sabwi- entry and the oi- entry both show “derivation: -i-”.
• The following “related:” field indicates related terms, not necessarily derived terms or derivation sources.

The oi- entry indicates that noun oya is related to the verb oi-, and also does not state that oya is derived from oi-, nor that oi- is derived from oya.

The sabwi- entry indicates that suffix -sabwi- is related to the verb sabwi-, and also does not state that -sabwi- is derived from sabwi-, nor that sabwi- is derived from -sabwi-.

• You are correct that there is no particular reason to exclude oi- here. However, no one has made any comment about excluding oi- from anything.

‑‑ Eiríkr Útlendi │^{Tala við mig} 01:25, 3 February 2023 (UTC)

Indeed, for words for which the ONCOJ dictionary gives a full analysis, e.g. akesar or aswobap, it uses different fields. But in general it does not attempt to provide full etymological information. Often there is just an affix without indicating what base form it was attached to, and/or links to transitive, intransitive or other related forms without specifying the direction of derivation.

I would be against including oi- as an example, precisely because we don't have a source for the proposed derivation. Kanguole 10:06, 3 February 2023 (UTC)

Here are some examples of other cases, I hope I can clarify myself better:

目(ma) + i = 見(mai) UM

mai > mi¹ > mi

座(wa) + i = 居(wai) UM

wai > wi¹ > wi > i

參(maw[a]) + i = 參(maw[a]i) UB

maw[a]i > mawi¹ > mawi > mai

親(oya) + i = 老(oyai) UB

oyai > oi¹ > oi

臥(kɵya) + i = 展(kɵyai) UB

kɵyai > ko²i¹ > koi

in short

Ca + i = Cai UM/B

Cai > Ci¹ The Young Prussian (talk) 19:49, 6 February 2023 (UTC)

@The Young Prussian: This is certainly interesting, but there are a number of potential issues.

• Regarding the phonology, I have read elsewhere (possibly in Shibatani? can't find it at the moment...) that one theory to explain those nouns that have both 露出形 (when used as standalone nouns; ending in -e or -i) and 被覆形 (when used in compounds; ending in -a, -o, or -u) is that there may have been fusion with a particle い, which does appear in very old texts as an emphatic nominalization particle. If this fusion theory is correct, then 目 (ma) + い (i) = 目 (me₂), not 見 (mi₁-).
• Also on the phonology, I'm not aware of any phoneme ⟨wi₁⟩, as you have listed in your reconstructions for both ゐる and まゐる. Considering that vowel sound ⟨i₁⟩ is often reconstructed as haveing a palatal glide to start, perhaps like /ji/, while ⟨i₂⟩ would have begun with a labiel glide like /wi/, no phoneme ⟨wi₁⟩ would have been possible.
• Regarding the proposed roots above, several of these seem ... odd. I am not familiar with 座 (wa; usually read as ざ or すわ-る), 参 (mawa; no such morpheme), 臥 (keya; usually read as が, or as ふ-せる or ふせ-る).

• Do you have any evidence for the existence of suggested root /wa/? And why do you spell that with the kanji 座? What is this root noun supposed to be?
• Do you have any evidence for the existence of suggested root /keya/? And why do you spell that with the kanji 臥? The only term match I'm finding is the Old Japanese -na adjective けや, apparently meaning "clearly; completely" when used adverbially, the only form for which I can find a citation. The semantic shift from "clearly; completely" to "to love" seems very strange indeed. Then the phonological shift from ⟨keyai⟩ to ⟨ko₂i₁⟩ also seems very odd: how do you explain the /e/ → /o/ change? I don't think I've ever seen this particular vowel change suggested anywhere else. What is this verb supposed to be? If you are referring to 恋いる (koiru), the historical kana spelling is こひる, and further digging reveals that this is a kami ichidan shift from older kopu, not koyu. This cannot derive from any purported root keya.
• Do you have any evidence for the existence of root /mawa/? One etymological theory posits that まゐる is the intransitive pair to まをす (i.e. 申す), much like the pair すぎる and すごす (Eastern Old Japanese in the Man'yōshū, compare Western すぐす), and that the base predicative form would have been /mawu/, realized in kana as まう. Alternatively, this may be derived from some prefixing element ま- + ゐる. Apparently this is only ever found in the 連用形 in ancient sources, making it impossible to tell anything further.

• Considering the existence of other ancient nouns ending in -a that appear to be closely related to verbs that manifest the yodan conjugation paradigm, I don't think we can make the kind of claim shown above, that -a nouns + -i produce kami ichidan (見る, 居る, 参る) and kami nidan (老ゆ) verb stems.

HTH, ‑‑ Eiríkr Útlendi │^{Tala við mig} 05:42, 7 February 2023 (UTC)

This is because the "i" found in "me" is a nominative particle, while in "mi" it is a verbalizing particle.

Logically, a + i should generate e², but since me and mi would share the same phoneme with different derivations, with the evolution from Late Proto-Japanese to Old Japanese mai (eye) and mai (see) gained tonal differences, and this is one of the reasons "e" became "i¹" in the transition from Late Proto-Japanese to Old Japanese.

"wi¹" does not actually exist, but that was just to specify the derivation of there is "i¹" standalone however it was for the same reason, there actually is no tonal variation preceded by a semivowel "w" and I didn't intend to change that, wi¹ actually represents "wai", as can also be found in "awau" (blue), but awau was derived from awo and not awo¹.

"kɵya" (not keya) is a verbal stem, koya is a derivation from koyasu and koyaru and both use the kanji 臥 as in MYS - 354, according to ONCOJ, whereas "wa" is a derivation from man'yougana in that an excerpt from Kojiki (or from the old part of MYS) 座 is used to represent "I/we" but I couldn't recover the excerpt, but I believe I found at least 3 excerpts where 座 is read as "wa".

In addition, in Old Japanese there was a strange type of onbin called オ音 (o-on) in which in certain verbs the contraction a + a would generate o, what happens with mawosu, which has mawi as a transitivity pair, would be the contraction from maw[a] plus the asu trasitivizer, the contraction a + a would generate o-on.

Furthermore, this parallel is possible due to the fact that they are all radical and abstract words, ending in -a, which is not restricted only to yodan, but also ku adjectives. As found in Vovin's Origins of the Japanese Language, where kusa-u, sira-u, awa-u, kuma-u and kura-u would have the same derivation, which would be a stem ending in a plus a nominative particle (similar to i) that would be derived from a common nominative particle to both Proto-Japanese and Proto-Korean "or". This derivation should therefore generate stem-type inflectional lemmas ending in -a + suffix. Such as -ku/-nari adjectives with the genitive meaning, -mu/-aru verbs with the "become" meaning, -aru/-ari verbs with the intransitive and stative meaning, -asu verbs with honorific and transitive meaning, -ziku adjectives with the semblative meaning and even -re demonstratives, so not only yodan arise from stems ending in a, actually, they are not even the majority.

And the koi is... [4]

And the mawi is UB in... [5]

Finally (I think), the derivation of koya, oya, wa and mawa is not different from kusa, kura, awa, kuma and sira:

proto-japonic

• in sira we have sira + u(or), sira + mu, sira + nari
• in kuma we have kuma + u(or)
• in awa we have awa + u(or)
• in kura we have kura + u(or), kura + si
• in kusa we have kusa + u(or), kusa + si, kusa + aru.
• in kɵya we have kɵya + i, kɵya + asu, kɵya + aru.
• in mawa we have mawa + i, mawa + asu (with o-on)
• in oya we have oya + i, oya + ⁿsi
• in wa we have wa + i, wa + ari (with o-on), wa + re

old japanese

• siro¹ (white), shiramu (to become white), siranari (whitten)
• kumo¹ (cloud)
• awo (blue)
• kuro¹ (black), kurasi (dark)
• kuso¹ (shit), kusasi (shity), kusaru (to become shit)
• ko²i (lie), ko²yasu (lie), ko²yaru (to be lying)
• mawi (visit, go), mawosu (speak)
• oi (to become old), oyazi (the same of)
• wi (sit), wori (to be siting), ware (we sitting around)

The derivation is the same, in all the fusion of vowels a + u/i generates diphthongs according to the resulting gramatical class:

Ca + u [nominalizer] = Co¹

Ca + i [verbalizer] = Ci¹

what happens to the yodan verbs is not different

Ca [irrealis] + i [verbalizer] = Ci¹ [infinitive]

this is the reason why the infinitive of yodan verbs ends in i¹ and the irrealis end in a.

However this derivation may present irregularities, since a + i can generate i¹ or e², in the same way, o² + i can generate i² or e² when the same particle is added to nominal stems ending in a/u/o² generating ichidan/nidan verbs, like sime² and oi, and in certain cases, yodan, like tuki¹, napi¹.

The Young Prussian (talk) 12:49, 7 February 2023 (UTC)

@The Young Prussian: Various issues remain.

• "This is because the "i" found in "me" is a nominative particle, while in "mi" it is a verbalizing particle."

You're describing grammar, not phonology. This does not explain why ma + i would become mi₁ instead of the expected me₂. Furthermore, this reads as your own theorizing, which we cannot use here at Wikipedia; are there any published authors arguing along similar lines? Any specific references you can point us to?

• "...wi¹ actually represents "wai", as can also be found in "awau" (blue), but awau was derived from awo and not awo¹."

This seems confused. The notation ⟨i₁⟩ is used in academic texts about Old Japanese to denote a particular variant of /i/ that is often reconstructed with a palatal glide, contrasting with ⟨i₂⟩ that is often shown with a labial glide. Any morpheme /wai/ could not phonologically manifest as ⟨wi₁⟩. This is simply phonologically impossible, from what I've read, and no reconstruction of Old Japanese that I have yet encountered differentiates the /i/ or /o/ vowels when preceded by /w-/.

Separately, where do you get awau? I have never encountered this reconstruction before.

• "“kɵya” (not keya) is a verbal stem..."

So you derive a verb... from a verb? Towards the top of this thread, you were arguing that kami verbs derived from nouns ending in -a. If kɵya is not an independent morpheme, then it is irrelevant to this discussion.

Separately, where do you get [ɵ]? I have not seen this used in other reconstructions of Old Japanese and Proto-Japonic, which instead use [ə] or [o] for the vowel value of ⟨o₂⟩.

• "whereas “wa” is a derivation from man'yougana in that an excerpt from Kojiki (or from the old part of MYS) 座 is used to represent "I/we" but I couldn't recover the excerpt, but I believe I found at least 3 excerpts where 座 is read as “wa”."

If you can share your specific source, that would be helpful.

The best I can find is 御座す (owasu, honorific for "to be; to go; to come"). However, this is from older kana spelling おはす, ruling out any wa origins. Moreover, this verb doesn't appear until the late 800s, and the KDJ entry suggests that this is a phonological contraction or shift from either おほまします → おはします → おはす, or from おほます → おはす.

• "In addition, in Old Japanese there was a strange type of onbin called オ音 (o-on) in which in certain verbs the contraction a + a would generate o, what happens with mawosu, which has mawi as a transitivity pair, would be the contraction from maw[a] plus the asu trasitivizer, the contraction a + a would generate o-on."

This is also new to me, and it strikes me as unlikely. There are other kami nidan verb pairs where the transitive ends in -su that appear to break that pattern, such as おつ → おちる・おとす, すぐ → すぎる・すぐす. There are also cases of yodan verbs with causative / honorific forms ending in -osu, such as おもふ → おもほす, contrasting with other yodan verbs with causative / honorific forms ending in -asu, such as とる → とらす. There are also yodan verbs with apparent derived forms that include both -aru and -oru endings, such as こむ → こまる・こめる・こもる, or つむ → つまる・つめる・つもる. There are also cases of nouns ending in -a that have related verbs ending in -aru or -asu, but not -oru or -osu, such as あか → あかる・あかす. If the causative / transitive / honorific ending is actually -asu (not just -su), and if a + a = o, then あか + あす = *あこす. But no such form exists for this verb. And then there's your putative verbal stem kɵya. The recorded honorific verb form koyasu would presumably be this stem kɵya + causative / transitive / honorific ending -asu. If the オ音 shift you describe is accurate, this should produce *koyosu, but again no such form exists for this verb.

Given the distinct possibility that mawiru and mawosu are not in fact cognate, and that mawiru may be ma- + wiru, and given also that the only form of mawiru that appears in OJP texts is the continuative stem form mawi-, I think we need other examples before positing such a phonological rule, especially considering the apparent exceptions.

The JA Wikipedia has no ja:オ音 or ja:オ音便 articles, and the extant ja:音便 article doesn't include the string "オ音" anywhere. I've tried googling for this "o-on", but that too was to no avail. Who else is writing about this オ音? What reference titles can you give us?

• "...this parallel is possible due to the fact that they are all radical and abstract words, ending in -a,..."

I'm afraid that your text here makes no sense to me. Your earlier example oya ("parent") is neither "radical" nor "abstract".  ??? Your later example kɵya does not even appear to be a word. Etc.

• "As found in Vovin's Origins of the Japanese Language, where kusa-u, sira-u, awa-u, kuma-u and kura-u would have the same derivation, which would be a stem ending in a plus a nominative particle (similar to i) that would be derived from a common nominative particle to both Proto-Japanese and Proto-Korean “or”."

There are no such words kusau, sirau, etc.  ??? Poking around in ONCOJ for 黒, I find more instances of ⟨kuro₁⟩ → /kurwo/ than of ⟨kuro₂⟩, and only ⟨kuro₂⟩ could conceivably derive phonologically from /kurau/. Similarly for 白, this appears as ⟨siro₁⟩ → /sirwo/, not as ⟨siro₂⟩. Guessing that your "kuma-u" is supposed to be 雲, OJP forms seem to show ⟨kumo₁⟩ → /kumwo/, again ruling out any /kumau/ phonological source.

Also, there is no need for any such putative nominative particle u: kura and oya are clearly independent nouns unto themselves.

• "Such as -ku/-nari adjectives with the genitive meaning..."

What "genitive meaning"?

• "...-mu/-aru verbs with the "become" meaning..."

So far as I'm aware, the -mu suffix and derivatives have a core meaning of "to look like, seem like, behave like", not "to become". I dimly recall reading that this might be related to me "eye" and miru "to see".

• "...-ziku adjectives with the semblative meaning..."

There are no real "-ziku adjectives".  ??? Morphologically, anything ending in -ziku would be an adverb, not an adjective. The only -ziku adjective I can find is 時じく in the MYS. The KDJ entry here describes this as a shift in usage of the 連用形 / adverbial form of older adjective 時じ (tokiji), and that entry explains this as noun 時 (toki, "time") + suffix ～じ (-zi, "-like", makes adjectives of nouns).

• "And the mawi is UB in... [6]"

I didn't ask about mawi. I asked for evidence of any morpheme mawa.

• "... so not only yodan arise from stems ending in a, actually, they are not even the majority."

I never made any such claim that yodan verbs "arise from stems ending in a". I do not think that yodan verbs arise from such stems. I don't think Kanguole made any such contention either. Moreover, if we did posit that yodan verbs arose from nouns ending in -a, how would the suffixation with some verb-forming element -i possibly work?

• "* oi (to become old), oyazi (the same of) // wi (sit), wori (to be siting), ware (we sitting around)"

Your semantics are very strange. "to become old" → "parent / elder" → "the same of"??? "sit" → "be sitting" → "we"???

The form おやじ for おなじ might well represent a pronunciation shift: we don't know which is older. The meaning of oya makes any root relationship deriving oyaji ↔ onaji unlikely at best.

The term ware is first-person singular at its root. Moreover, the core morpheme is just wa, apparently related to alternative form a, as we see in phrasing like わぎも, わがえ, あがきみ, etc. The meaning of wa / a / ware makes any derivation from wi extremely unlikely.

• "The derivation is the same, in all the fusion of vowels a + u/i generates diphthongs according to the resulting gramatical class: // Ca + u [nominalizer] = Co¹ // Ca + i [verbalizer] = Ci¹"

As described above, the phonology is wrong. a + u should result in ⟨o₂⟩, not ⟨o₁⟩. Likewise, a + i should result in ⟨e₂⟩, not ⟨i₁⟩. Even if we allow for fronting all the way to /i/, we would expect ⟨i₂⟩, not ⟨i₁⟩.

• "what happens to the yodan verbs is not different // Ca [irrealis] + i [verbalizer] = Ci¹ [infinitive] // this is the reason why the infinitive of yodan verbs ends in i¹ and the irrealis end in a."

Your argument here is circular -- you're basically saying that the yodan verb irrealis ends in -a because the yodan verb irrealis ends in -a.

I'm afraid that your post is hard to follow, and there are many apparent problems in what you have to say. In addition, without any references, I can only conclude that these are your own thoughts, which we cannot use for purposes of Wikipedia articles. As it is, I have spent far too much time on this. I wish you luck and clarity. ‑‑ Eiríkr Útlendi │^{Tala við mig} 23:27, 7 February 2023 (UTC)

Central/Western Old Japanese

This is responding to this discussion about what to call the OJ of the Nara area (which reflects the vast bulk of what is attested) in contrast to Eastern Old Japanese.

• Frellesvig speaks of "central OJ", Unger of "(central) OJ", i.e. not a proper noun.
• Several authors (e.g. Miyake, Bentley, de Boer) call it "Central OJ".
• Vovin (who is very prolific) calls it "Western OJ", presumably to give it equal status with the much-less-attested EOJ. (But he says it is part of Central Japonic.) Vovin's student Kupchik uses the same term, and even introduces "Central OJ" for Sinano dialects of OJ, which other authors include in EOJ.

Kanguole 19:13, 1 December 2022 (UTC)

I₁

In fact, it is recognized here that there is not yet a division between the standalone vowels i₁ and i₂, however in the Kojiki, this type of division seems apparent, which also already had a division between mo₁ and mo₂. Because some transliterations put i₁ma/i₁me (dream) as jima/jime, the same with i₁ (sleep) as ji, they all use man'yougana 伊 to represent this supposed ji, which is also used to represent the stem form of the verb oyu (to age) which would be oi₁, but in this case it is oji, which makes sense because they are in the same column, could someone clarify if there really is such a difference and if the syllable ji exists in old Japanese.

Source

https://oncoj.orinst.ox.ac.uk/cgi-bin/oncoj_tgrep2.sh?search=/L050103/

https://oncoj.orinst.ox.ac.uk/cgi-bin/oncoj.sh?tree=KK.93@78

https://en.wiktionary.org/wiki/%E5%AF%9D#Etymology_3

--The Young Prussian (talk) 01:50, 22 December 2022 (UTC)

palatalization before /e/

we say

The sibilants /s/ and /ⁿz/ may have been palatalized before e and i

Which /e/? I assume if it is only one, it is the /e/ that comes from earlier /ia/. Or is it both? —Soap— 21:33, 24 August 2023 (UTC)

The statement (from Miyake) relates to the OJ period, when (according to the script) there was only one /e/ after /s/ and /ⁿz/. That /e/ was probably the result of a merger of two or more vowel sequences from a pre-OJ period, but in OJ they had already merged, and Miyake is talking about allophones of /s/ and /ⁿz/ before that single /e/. Kanguole 22:00, 24 August 2023 (UTC)

This could be taken as asking whether e₁ and e₂ merged as e₁ or e₂ after /s/ and /ⁿz/. Miyake (p264) believes it was the former, but others have different views. Kanguole 21:15, 25 August 2023 (UTC)

Okay thanks. I think I understand now. The shifts /s z/ > /š ž/ and the opposite are both about equally likely, so we can't just look at morphology and assume that if we see underlying /sia/ that it must have been /še/. And the kanji were already merged in even the earliest texts, so we can't look at Chinese either. —Soap— 14:50, 6 September 2023 (UTC)

FWIW, while quite a bit later, the 1603 Nippo Jisho records that せ (modern se) was pronounced at that time as something more like /ʃe/. ‑‑ Eiríkr Útlendi │^{Tala við mig} 23:34, 24 January 2024 (UTC)